Absolute pitch judgments of familiar melodies generalize across timbre and octave

Abstract

Most listeners can determine when a familiar recording of music has been shifted in musical key by as little as one semitone (e.g., from B to C major). However, it is unclear how this form of pitch memory relates to absolute pitch (AP) representations, which are based on pitch chroma. This is because listeners could use spectral cues from the familiar instrumentation of the original recordings or strategies based on pitch height (e.g., relying on a feeling that an incorrect recording sounds “too high” or “too low”) to determine when a familiar recording has been shifted in pitch. Neither of these strategies would require the kind of understanding of pitch chroma or musical key associated with AP. The present experiments thus assessed whether listeners could make accurate absolute pitch judgments when listening to novel renditions of these familiar melodies, differing from the iconic recording in complexity and timbre (Experiment 1) or timbre and octave (Experiment 2). These experiments eliminate the possibility of reliance on spectral cues and pitch height, respectively. Listeners in both experiments selected the correct-key version of the familiar melody at rates that were significantly above chance. These results fit within a growing body of research supporting the idea that most listeners, regardless of formal musical training, have robust representations of absolute pitch - based on pitch chroma - that generalize to novel listening situations. Implications for theories of auditory pitch memory are discussed.